Unlock instant, AI-driven research and patent intelligence for your innovation.

Management method, device and system for computational nodes

A technology of computing nodes and management methods, applied in the field of network communication, which can solve problems such as unproposed solutions, waste of energy and computing resources, and low load occupation

Active Publication Date: 2013-05-01
BEIJING RUNSTONE TECH
View PDF5 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] Large-scale IT operation enterprises and operators have tens of thousands of devices running on the network. Since the operating system, specific applications, and application levels are all built and operated independently, the load occupation of each system in daily operation is generally relatively low, and many even occupy less than More than 10%, in order to ensure the high availability and performance of the system, energy-saving deployment is generally not done by turning off the CPU, reducing the main frequency, and turning off unused hard disks; in a controllable scale network, there are a lot of loose computing resources , which results in unnecessary waste of energy and computing resources
[0003] For the above-mentioned problems in related technologies, no effective solution has been proposed yet

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Management method, device and system for computational nodes
  • Management method, device and system for computational nodes
  • Management method, device and system for computational nodes

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0019] figure 1 It is a flowchart of a computing node management method according to Embodiment 1 of the present invention. Such as figure 1 As shown, the method includes:

[0020] Step S102, the server in the cloud management center monitors the resource usage rules of each computing node;

[0021] Step S104, the above-mentioned server acquires a resource usage policy according to a resource usage rule, wherein the resource usage policy corresponds to a task type assigned to each computing node;

[0022] Step S106, the server sends the resource usage policy to each computing node.

[0023] In this embodiment, there are many ways to monitor the usage rules of the above-mentioned resources. For example, it can be monitored through a specific local program, or it can be realized through the probe downloaded from the above-mentioned server. For the latter processing method, the following process can be used specifically. : Before the server monitors the resource usage rules o...

Embodiment 2

[0041] This embodiment adopts Figure 5 The topology shown is realized.

[0042] 1. Definition of key content

[0043] 1.1 Weiyun probe (equivalent to the resource probe in Example 1)

[0044] The micro-cloud probe is used as the early resource detection of the micro-cloud system deployment. All systems can download their own corresponding probes. The initial probe is only used for resource registration and performance monitoring, and the monitoring results are uploaded to the micro-cloud core management center. After a long-term After monitoring and confirming the regularity of system resource usage or long-term stable idle resources, the micro-cloud center formulates resource usage policies and actively downloads them to the node machine for local execution by resource probes;

[0045] Micro-cloud probes can be divided into several operating levels (active upgrade or shutdown):

[0046] Primary - resource registration and monitoring type;

[0047] Intermediate - task pro...

Embodiment 3

[0067] Deployment example of micro cloud system

[0068] Distributed large log analysis deployment

[0069] Taking the domain name resolution log or radius log as an example, there are nearly 10 billion pieces of data every day. If you want to build a system for full-text retrieval and analysis of these data, you may need to build a powerful hardware cluster system and a complex software system. It may be completed, but the application and actual value generated by these data have not yet been discovered. It seems directly that the input-output ratio is not directly proportional. It can be realized by micro-cloud technology, and does not require a large-scale increase in hardware investment;

[0070] The log data to be analyzed is directly transmitted to the computing nodes according to the computing units and strategies allocated by the cloud control center. The cloud center provides computing models and analysis templates, issues query and analysis instructions, and summariz...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides management method, device and system for computational nodes, wherein the method comprises the steps of: monitoring the resource usage pattern of each computational node by a server of a cloud management center; acquiring resource usage strategies by the server according to the resource usage patterns, wherein the resource usage strategies correspond to a task type assigned to each computational node; and sending the resource usage strategies to each computational node by the server. According to the adoption of the technical scheme provided by the invention, the technical problems that unnecessary energy and computing resources waste exists in networks of a certain scale, and the like in the related technologies are solved, thus both cost and computing resources are taken into account, that is to say, idle computing resources are utilized reasonably while no energy is wasted.

Description

technical field [0001] The present invention relates to the field of network communication, in particular to a management method, device and system for computing nodes. Background technique [0002] Large-scale IT operation enterprises and operators have tens of thousands of devices running on the network. Since the operating system, specific applications, and application levels are all built and operated independently, the load occupation of each system in daily operation is generally relatively low, and many even occupy less than More than 10%, in order to ensure the high availability and performance of the system, energy-saving deployment is generally not done by turning off the CPU, reducing the main frequency, and turning off unused hard disks; in a controllable scale network, there are a lot of loose computing resources , which causes unnecessary waste of energy and computing resources. [0003] For the above-mentioned problems in related technologies, no effective so...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): H04L12/24H04L29/08
Inventor 王立俊杨世标黎宇
Owner BEIJING RUNSTONE TECH